Quote:
Originally Posted by RWebb
you are both right - if the price paid buys assets all of which are worth zip; then the cost is 700B
otherwise 700B - x
unlikely they all are worth nothing - esp. the land value
|
Here is something that might be interestng - scenarios for these illiquid assets' values.
Here's a very rough model (image pasted in below) for US subprime mortgages, by vintage (year the loan was written). Appx $390BN of subprime mortgages (measured by principal) was written in 2006. There is only appx $6BN that was written in 2001, partly because there was less subprime lending then and partly because of all the subsequent refinancing.
For each vintage, make an assumption for foreclosure rate (percent of homes that will ultimately be foreclosed on) and loss rate (when the foreclosed homes sell at auction, what percent of the loan principal the bank will lose). E.g. if the home sold in 2006 for $350K with no down payment, then the principal is still roughly $350K today, if the house sells at foreclosure auction for $200K net of costs, the bank loses $150K = $350K - $200K and the loss rate is 150/350 = 42%. Given those assumptions, you can estimate the "hold-to-maturity" fair value of all existing US subprime mortgages, using loss = principal x foreclosure rate x loss rate, and fair value = principal - loss.
What are some possible scenarios?
The "GOOD" scenario is roughly what JP Morgan is using to value the mortgages and similar assets it acquired with WaMu. Suggests subprime mortgages may be worth 80 cents on the dollar. I'm calling this a reasonable scenario, but you have to make your own judgment.
The "REALLY BAD" scenario is, what I'm calling the Great Depression scenario, but you can use an even worse scenario if you want. 80% of 2006-vintage subprimes are foreclosed, and the houses sell for 10% of their original price - so that formerly $350K house sells for $35K. Here, suggests subprime mortgages may be worth 40 cents on the dollar.
In the in-between "BAD" scenario, suggests 65-70 cents.
By the way, I focus on the 2006 vintage, which is the "worst" as that was the peak of the housing bubble. Then I assumed earlier vintages get gradually "better" (lower foreclosure and loss rates). But, since the bulk of existing subprime mortgages were written in 2004-2007, shouldn't make that much difference how the earlier vintages perform.
Right now, I've read some senior subprime MBS are quoted around 60 cents on the dollar, although hardly any are actually trading. Possibly banks may have marked those assets to 85-95 cents on the dollar. If a desperate bank had to sell a large quantity of such assets into the frozen markets right now, it would take a huge bath, might only get 20-30 cents on the dollar. Suppose govt buys the asset for, let's say, 90 cents. And suppose the "BAD" scenario turns out to be correct, and the govt eventually recovers 65 cents for the asset for a net cost/loss of 35 cents. Scale that up to $700BN of assets purchased, then you'd get about $195BN of ultimate net cost/loss for the govt.
No-one knows for sure what the outcome will be, or exactly what assets will be purchased, but I find it interesting to think about the range of potential outcomes. Your conclusion will depend on your own assumptions of course. Maybe I am too optimistic, or too pessimistic.
Disclaimer - not investment advice. Simply some "thinking out loud".